Artificial intelligence in the diagnosis of dental diseases on panoramic radiographs: a preliminary study

BMC Oral Health

Table 2 Diagnostic performance of the framework and individual dentists for impacted teeth

Var	Spe	Sen	Youden’s index	AUC (SE; 95% CI)	P	Z
Framework	0.996	0.964	0.960	0.980 (0.00480; 0.976–0.983)	ref	ref
H1	0.997	0.987	0.984	0.992 (0.00291; 0.990–0.994)	0.009*	2.629
H2	0.998	0.966	0.964	0.982 (0.00463; 0.979–0.985)	0.607	0.514
H3	0.997	0.956	0.953	0.976 (0.00527; 0.973–0.980)	0.501	0.673
M1	0.998	0.924	0.922	0.961 (0.00676; 0.957–0.966)	0.013*	2.487
M2	0.999	0.932	0.931	0.966 (0.00642; 0.962–0.970)	0.049*	1.972
M3	0.997	0.964	0.961	0.980 (0.00480; 0.977–0.983)	0.906	0.118
L1	0.999	0.878	0.877	0.938 (0.00837; 0.933–0.944)	< 0.0001*	5.009
L2	0.999	0.901	0.900	0.950 (0.00763; 0.945–0.955)	0.000*	3.876
L3	1.000	0.883	0.883	0.941 (0.00822; 0.936–0.947)	< 0.0001*	4.531

Delong test was conducted for statistically analysis of the difference of AUC between the framework and the dentists. * represents significant difference (p < 0.05)

ISSN: 1472-6831